Multiple Reaction Monitoring-Ion Pair Finder: A ... - ACS Publications

Subscriber access provided by NEW YORK UNIV

Technical Note

MRM-Ion Pair Finder: a systematic approach to transform nontargeted mode to pseudo-targeted mode for metabolomics study based on liquid chromatography-mass spectrometry Ping Luo, Weidong Dai, Peiyuan Yin, Zhongda Zeng, Hongwei Kong, Lina Zhou, Xiaolin Wang, Shili Chen, Xin Lu, and Guowang Xu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b00615 • Publication Date (Web): 17 Apr 2015 Downloaded from http://pubs.acs.org on April 21, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

MRM-Ion Pair Finder: a systematic approach to transform non-targeted mode to pseudo-targeted mode for metabolomics study based on liquid chromatography-mass spectrometry Ping Luo♠, Weidong Dai♠, Peiyuan Yin, Zhongda Zeng, Hongwei Kong, Lina Zhou, Xiaolin Wang, Shili Chen, Xin Lu, Guowang Xu*

Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China

♠

: equal contribution.

*

: Corresponding author:

Prof. Dr. Guowang Xu, Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China. Tel./Fax: +86-411-84379530. E-mail: [email protected].

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

ABSTRACT: Pseudo-targeted metabolic profiling is a novel strategy combining the advantages of both targeted and untargeted methods. The strategy obtains metabolites and their product ions from Q-TOF MS by information-dependent acquisition (IDA), then picks targeted ion-pair and measures them on a triple-quadrupole MS by multiple reaction monitoring (MRM). The picking of ion pairs from thousands of candidates is the most time-consuming step of pseudo-targeted strategy. Herein, a systematic and automated approach and software (MRM-Ion Pair Finder) were developed to acquire characteristic MRM ion pairs by precursor ions alignment, MS2 spectrum extraction and reduction, characteristic product ion selection, and ion fusion. To test the reliability of the approach, a mixture of 15 metabolite standards was firstly analyzed and the representative ion pairs were correctly picked out, then pooled serum samples were further studied and the results were confirmed by the manual selection. Finally, a comparison with a commercial peak alignment software was performed, a good characteristic ion coverage of metabolites was obtained. As a proof of concept, the proposed approach was applied to a metabolomics study of liver cancer, 854 metabolite ion pairs were defined in the positive ion mode from serum. Our approach provides a high throughput method which is reliable to acquire MRM ion pairs for pseudo-targeted metabolomics with improved metabolite coverage and facilitate more reliable biomarkers discoveries.

2


Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Metabolomics, defined as the measurement of endogenetic metabolites in a given system, is playing substantial roles in (patho) physiological mechanism investigations and biomarker discovery studies.1-3 Non-targeted analysis is the most commonly used strategy

in

the

liquid

chromatography-mass

spectrometry

(LC-MS)-based

metabolomics studies.4 With high-resolution mass spectrometer employed, such as time-of-flight (TOF), Orbitrap, and Fourier transform ion cyclotron resonance MS, non-targeted methods can detect many metabolites in a non-prejudiced mode and provide accurate mass to facilitate the compound identification.5 However, non-targeted method is queried for its disadvantages in the quantification of the metabolome including limited repeatability and linear range, and the complicated peak picking and alignment algorithms which unavoidably introduce some false results.6,7 Multiple reaction monitoring (MRM) selecting characteristic precursor ion and product ion to measure known metabolite,8-11 which is often performed on triple quadrupole (TQ) mass spectrometer, is considered as the gold standard for compound quantification.12 This targeted method provides improved repeatability and linear range (4 - 5 orders of magnitude),13 and avoids the false results arising from peak alignment process since only limited number of known metabolites are measured, this is also the major disadvantage of MRM method.14 To date multiple targeted metabolomics methods have been applied for multiple metabolite analysis, but the targets are still limited to the known metabolites and are powerless for the unknown metabolites.15-17 To resolve these questions, we proposed a novel approach combining the advantage of both comprehensive coverage (full scan) and accurate quantification (MRM) methods, which was defined as pseudo-targeted 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 24

metabolomics method.18 It consists of three steps including non-targeted profiling on Q-TOF MS, ion-pair picking and MRM measurement on TQ MS. Nontargeted profiling provides rich information of metabolic ions based on the MS priority of high sensitivity, high resolution and high throughput. After picking out the targeted ion pairs, these metabolic ions could be quantified on TQ MS with better quality. In contrast with multiple targeted metabolomics, although metabolites are not identified, a pseudo-targeted method can provide global metabolome information with the aid of real samples.19,20 The key of this method is to get the characteristic ion pairs from tens of thousands auto-MS2 spectra scanned from a real biological matrix (e.g., plasma, serum, and urine). Although current commercial softwares may be used to handle the MS raw data,21-23 to the best of our knowledge, there is no software tailored to get the MRM transition lists for representing the metabolic information of real biological samples in the metabolomics study. Therefore, an automated and reliable approach is in a urgent need to accelerate this key process and to improve the data quality of global metabolome measurements. In this work, we presented a systematic procedure to scan and find out candidate ion pairs for pseudo-targeted metabolomics. A homemade software, MRM-Ion Pair Finder, was developed for high throughput selection of characteristic MRM ion pairs and

their

corresponding

collision

energy

(CE)

voltages

generated

from

information-dependent acquisition (IDA)-based auto-MS2 of biological samples. As a proof of concept, the proposed method was applied to the serum pseudo-targeted metabolomics study of liver cancer.

4


Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


EXPERIMENTAL SECTION Sample collection and pretreatment. The fasting serum samples of 30 healthy controls and 30 hepatocellular carcinoma (HCC) patients were collected from the First Hospital of Jilin University (Changchun, China). The transportation and storage of the samples were according to the standard procedures of metabolomics studies.24 Two pooled samples were prepared by mixing the aliquots of each sample in the healthy control group and the HCC group, respectively. A 200 µL acetonitrile was added into 50 µL pooled sample for the protein precipitation, then, the sample was vortexed for 60 s and then centrifuged (14,000 rpm, Biofuge Stratos, Thermo Scientific, USA) for 10 min at 4 °C. A portion of 200 µL supernatant was transferred to an Eppendorf tube for lyophilizing. The residue was reconstructed by 50 µL solution (acetonitrile/water = 1/3, v/v). The supernatant was used for the LC-MS analysis. LC-MS analysis. The untargeted analysis was performed on the Acquity ultra performance liquid chromatography (UPLC) system (Waters, Milford, MA, USA) coupled with a Triple TOF 5600 mass spectrometer system (AB SCIEX, Framingham, USA), operated in the positive electrospray ionization (ESI+) mode. The chromatography separation was carried out on a Waters Acquity BEH C8 column (100 mm × 2.1 mm, 1.7 µm). The mobile phase A was 0.1% (v/v) formic acid aqueous solution and the phase B was 0.1% (v/v) formic acid acetonitrile solution. The LC program begun with 10% B, linearly increased to 40% B at 3 min then to 100% B at 15 min and held for 5 min. Another 3 min was assigned for equilibration before next analysis. The injection volume was 5 µL. The flow rate was 0.35 mL/min and the column temperature was set at 50 °C. For the mass spectrometer system, the 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 24

curtain gas, gas1 and gas2 were 0.241, 0.276 and 0.276 MPa, respectively. The temperature of electrospray ion source was set at 550 °C. The voltage of spray was set at 5,500 V. The IDA-based auto-MS2 was performed on the 20 most intense metabolite ions in a cycle of full scan (0.25 s). The scan range of m/z of precursor ion and product ion were set as 100 − 1,200 Da and 80 − 1,000 Da, respectively. The CE voltage was set at 20, 40, and 60 V in the positive ESI mode. MRM ion pair selection. The MRM ion-pair selection including precursor ion alignment, auto-MS2 spectra data extraction and reduction, characteristic product ion selection and ion fusion were implemented by the homemade software MRM-Ion Pair Finder. According to our previous study,25 ion fusion was performed to reduce redundant ions which originate from the same metabolites, such as the isotopic ions, adduct ions and fragment ions.26,27 The software tool, MRM-Ion Pair Finder, was performed by the in-house coded algorithms with the Matlab environment (Version 7.14.0.739, R2012a, 64-bit). Data conversions. Prior to the operation of MRM-Ion Pair Founder software raw data were converted into a specific format. Two file folders were needed for the conversion, one folder includes the lists of precursor ion information from multiple LC-MS runs at different CE voltage, and the other contains the product ions information. Raw data files (WIFF format) of IDA-based auto-MS2 analysis at CE voltage of 20 V, 40 V and 60 V were firstly converted to MGF files by PeakView software V1.2.0.3 (AB SCIEX, USA) and then converted to TXT files by Notepad (Microsoft, USA), which include information of mass-to-charge ratio (m/z), retention 6


Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


time (tR), charge, and intensity of product ions. And the lists of corresponding precursor ions, including m/z, tR and intensity of precursor ion were exported from the IDA table in the PeakView software in the TXT form.

RESULTS AND DISCUSSION Overview of the procedure of metabolite MRM ion pair selection. A typical workflow of LC-MS-based pseudo-targeted metabolomics analysis and MRM ion pairs selection is illustrated in Figure 1. IDA-based auto-MS2 experiments with various CE voltages (20 V, 40 V and 60 V) of a pooled sample were firstly performed on a Triple TOF 5600+ MS to acquire high-resolution mass spectra of metabolite precursor ions and generated product ions. Then, MRM-Ion Pair Finder software was applied to select the most intensive product ion of precursor ion as the MRM product ion and the corresponding fragment voltage as optimal CE voltage to construct MRM transitions list. Next, Ions Fusion was applied to exclude redundant MRM ion pairs (originated from isotopic ions, adduct ions and fragment ions) from the MRM transitions list. Finally, a pseudo-targeted metabolomics analysis of batch samples in the dynamic MRM mode was performed on a Q-Trap 5500 mass spectrometer. Key steps of MRM ion pair selection including precursor ions alignment, auto-MS2 information extraction and reduction, characteristic product ion and CE voltage selection, as well as the following step of ion fusion are described as follows in details. Precursor ion alignment: The alignment began with the removal of noise ions according to user-defined noise level and ratio of signal-to-noise. Then precursor ions 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 24

from multiple runs were aligned, and shift of tR, tolerance of m/z and numbers of occurrence were adjusted according to the experiment (e.g., 0.3 min, 0.01 Da and 2, respectively). The result of alignment was output in a csv file ‘A’. Then, the TXT files containing corresponding auto-MS2 information were imported to the software to find the product ions for each precursor ion listed in the result. Auto-MS2 information extraction and reduction: Each IDA-based auto-MS2 analysis of a pooled serum sample resulted in ~28,000 MS2 spectra and an average number of 12 product ions in each MS2 spectrum. However, many of them were redundant due to coming from the same metabolites. Hence, it was needed to extract and simplify the auto-MS2 information before selecting characteristic product ions. We suggest 2 criteria for product ions: 1) [m/z

(precursor ion)

-m/z

(product ion)]

> 13.9 Da,

which means at least 14 Da (-CH2- group) is lost; and 2) the intensity of the product ion is more than 200 counts (the threshold of noise). In addition, for each of above mentioned aligned precursor ions, at most three of the most intensive product ions from each MS2 file were listed in a csv file ‘B’, and if no product ions were remained, the precursor ion was defined as its product ion at low CE voltage (e.g., 10 V). Characteristic product ion selection: After auto-MS2 information extraction and reduction, the most intensive product ion is to be selected as the most suitable product ion of the precursor ion, and the corresponding CE voltage is used as the optimal fragmentor of the MRM ion pair. Then, a MRM ion pairs list (CSV. format) including the information of precursor ion, fragment ion, CE voltage and tR was exported in a csv file ‘C’. This step was also performed using MRM-Ion Pair Finder software.

8


Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ions fusion: MRM ion pairs from isotopic ions, adduct ions (e.g., [M+Na]+, [M+K]+ or [M+NH4]+), fragment ions (e.g., loss of H2O, CO2), and oligomers (2M+H, 3M+H) may also emerge in the MRM transitions list and share scan time. Therefore, an ion fusion step was included to delete these redundant MRM transitions. Finally, a reduced MRM transitions list, saved in a csv file ‘D’, was directly imported into a Q-trap 5500 mass spectrometer system for dynamic MRM analysis of batch samples. Evaluation of reliability. To evaluate the feasibility of our proposed ion pair selection procedure, 15 single standard solutions (see Table S1, Supporting Information) were analyzed by the IDA-based auto-MS2 method. They were analyzed at CE voltage of 20 V, 40 V and 60 V in the ESI+ mode, respectively. Their characteristic product ions and optimal CE voltages were correctly found (see Table S1, Supporting Information). Then, a mixture containing 15 metabolite standards was repeatedly analyzed three times at each CE voltage to evaluate the reliability of the software tool. After precursor ion alignment, auto-MS2 extraction and reduction, characteristic product ion and optimal CE voltage selection, and removal of background ions, 22 (all of them were related to 15 standards) ion pairs can be steadily found by MRM-Ion pair Finder software in 3 runs, and after the ions fusion, a list of MRM ion pairs only including all 15 standards with a representative form was generated (see Table S1, Supporting Information). These standard-based results were completely in conformity with the manual operation. The reliability was further evaluated by comparing the results obtained from newly developed software (MRM-Ion Pair Finder) and manual selection of pooled serum samples of healthy control group and HCC group at 3 CE voltages. After the precursor ion alignment, auto-MS2 extraction and reduction, characteristic product ion 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 24

and optimal CE voltage selection, and background ions elimination were performed, 1,446 MRM ion pairs were obtained (see Table S2, Supporting Information), and metabolites eluted within 10–12 min were exemplarily studied, 128 and 106 MRM ion pairs were obtained by MRM-Ion Pair Finder software and manual operation, respectively (see Table S3, Supporting Information). It is worth mentioning that the processes of manual operation, picking precursor ion with the same rules as the software MRM-Ion Pair Finder did, recording the most intensive suitable product ion of each picked precursor ion from every LC-MS run and then selecting the most intensive ions of them as the characteristic product ion of the precursor ion, was a time-consuming step and spent about 8 times longer time than software MRM-Ion Pair Finder did. Among these results, 106 precursor ions, 102 precursor ion-product ion pairs, and 102 precursor ion-product ion-CE voltage transitions were the same between two processing methods (Figure S1a, Supporting Information). A small number of different product ions and CE voltages were caused by different algorithms of product ion intensity between raw WIFF data (profile) and converted MGF data (centroid) (Figure S1b-d, Supporting Information). This situation occasionally occurs when two product ions have very close intensity. At this case, either of two product ions can be assigned as characteristic product ion. Of note, about 21% more MRM transitions (mainly from metabolites with low intensity) than manual selection were found, which makes a more comprehensive metabolite coverage possible, furthermore, the processing time was markedly shortened by the software. These results demonstrate that the proposed procedure is reliable and efficient to transform non-targeted full scan to pseudo-targeted metabolomics.

10


Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Evaluation of metabolite coverage. An ideal MRM transitions list of pseudo-targeted metabolomics should cover the metabolites in non-targeted full scan. The metabolite coverage of MRM transitions was compared by using 2 pooled serum samples with the full scan results processed by MarkerView V.1.2.1 (AB SCIEX, USA) which can be used to define the characteristic ions (not ion pairs), the peak picking and alignment software provided by the instrument vendor. In order to compare the results at the same conditions, noises were filtered with the noise level of 10000 units and ratio of signal to noise of 3 before peak picking of both methods. In the peak alignment both methods were operated with set of tR shift of 0.3 min, and m/z shift of 0.01Da. As shown in Figure 3a, full scan analysis with the MarkerView resulted in 1,389 metabolite features, while 1,446 metabolite MRM transitions were obtained by the MRM-Ion Pair Finder (see Table S2, Supporting Information). Among them, 1,317 metabolite features were the same (Figure 2b), and were covered by the two methods (Figure 2c, 2d). Only 72 and 129 were unique for each method. After carefully checking the mass spectra, they were found to be characteristic ions. The reason why two methods lost some ions may be the difference of algorithm or the input data forms (profile or centroid). The MRM-Ion Pair Finder can fuse features from the same metabolite. The redundant ions, such as isotopic ions, adduct ions and fragment ions, were excluded by the following ion fusion step, a tidy list of 854 MRM transitions (see Table S4, Supporting Information) was finally obtained from pooled serum samples. Application of the pseudo-targeted method in a metabolomic study. As an example, our method was applied to a serum metabolomic study of hepatocellular carcinoma (HCC). The list of 854 MRM ion-pairs from positive mode (see Table S4, 11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 24

Supporting Information) was used for next dynamic MRM analysis of batch samples including 30 healthy controls and 30 HCC patients on a Q-Trap 5500 MS system. Typical chromatogram of serum sample of MRM analysis in positive mode was shown in the Figure S2. The QC samples which were prepared by mixing the equal volume aliquots of each sample in the healthy control group and the HCC group were clustered tightly together on the score plot of principal component analysis (PCA) analysis and obvious separation could be found from the score plot of partial least-squares discriminant analysis (PLS-DA) (see Figures S3, Supporting Information). Furthermore, relative standard deviations (RSD) of 91% contents of the analytes were less than 30% from QC samples in positive ESI mode (see Figure S4, Supporting Information). Metabolites were identified by searching online databases such as HMDB and Metlin, and literatures according to the MS/MS spectra, and checked by the authentic standards. The differential metabolites study revealed the deregulated metabolism of amino acids, lipids, bile acids, and acylcarnitines in patients with HCC (see Figure 3a), and the information of the top 20 differential metabolite ion-pairs is listed in the Table S5, it can be known that the changes of the identified metabolites were consistent with our previous study.19,28,29

30

Serum LPCs

significantly decreased in the HCC patients. LPCs are important biomolecules and play a vital role in the regulation of inflammation, cell proliferation and cell invasion,31 LPC 22:5 was reported as one of the potential biomarkers to distinguish HCC from other liver diseases.29 Bile acids, important signal metabolites of energy homeostasis,32 were remarkably elevated, which may be the result from abnormality of energy metabolism in the HCC patients.

12


Page 13 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In the pseudotargeted method, the characteristic ion pairs are from the real samples including the identified and unidentified metabolites. It is worth noting that some important unidentified metabolite ions were also detected with MRM method and traced back to the MS/MS information (see Figure 3b-d), and can be focused in the further study, which can widen the range of targets and facilitate the discovery of new important metabolites. For example, ions of No.749, No.119 and No.1354 (see Figure 3a), changed significantly in HCC, the pseudotargeted method provides us a chance to identify them. The metabolite with m/z 583.26 (No.1354) was further identified as biliverdin by using our previously developed method.33. In a word, the transition list obtained by the proposed approach can be successfully used to define the known and unknown metabolites with the advantages of a targeted method and a non-targeted method.

CONCLUSIONS Transforming non-targeted full scan to pseudo-targeted MRM analysis is a practicable way to improve the quality of metabolomics data and therefore the reliability of potential biomarkers. In this work, we developed an efficient approach for automated processing of IDA-based LC-MS/MS spectra to generate characteristic MRM ion pairs and optimal CE voltages. This method was implemented in a software tool, MRM-Ion Pair Finder. From the applications to metabolite standards and real biological samples, we demonstrated that MRM-Ion Pair Finder is a reliable tool for covering global metabolome by pseudo-targeted metabolomics method. 854 (ESI+ mode) metabolite MRM ion pairs were discovered in an exemplary application of 13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 24

human serum. We will continue to apply this approach to pseudo-targeted metabolomics studies of urine, plasma, tissue and other biological sample to facilitate more reliable biomarkers discovery.

ACKNOWLEDGMENT The study has been supported by the State Key Science & Technology Project for Infectious Diseases (2012ZX10002011), the foundation (No. 21375127) and the creative research group project (No. 21321064) from National Natural Science Foundation of China.

Supporting information available Additional information as noted in the text. This information is available free of charge via the Internet at http://pubs.acs.org/.

14


Page 15 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1) Szpunar, J. The Analyst 2005, 130, 442-465.

(2) Mapstone, M.; Cheema, A. K.; Fiandaca, M. S.; Zhong, X.; Mhyre, T. R.; MacArthur, L. H.; Hall, W. J.; Fisher, S. G.; Peterson, D. R.; Haley, J. M.; Nazar, M. D.; Rich, S. A.; Berlau, D. J.; Peltz, C. B.; Tan, M. T.; Kawas, C. H.; Federoff, H. J. Nat. Med. 2014.

(3) Sreekumar, A.; Poisson, L. M.; Rajendiran, T. M.; Khan, A. P.; Cao, Q.; Yu, J.; Laxman, B.; Mehra, R.; Lonigro, R. J.; Li, Y.; Nyati, M. K.; Ahsan, A.; Kalyana-Sundaram, S.; Han, B.; Cao, X.; Byun, J.; Omenn, G. S.; Ghosh, D.; Pennathur, S.; Alexander, D. C.; Berger, A.; Shuster, J. R.; Wei, J. T.; Varambally, S.; Beecher, C.; Chinnaiyan, A. M. Nature 2009, 457, 910-914.

(4) Theodoridis, G. A.; Gika, H. G.; Want, E. J.; Wilson, I. D. Anal. Chim. Acta. 2012, 711, 7-16.

(5) Dunn, W. B.; Broadhurst, D. I.; Atherton, H. J.; Goodacre, R.; Griffin, J. L. Chem. Soc. Rev. 2011, 40, 387-426.

(6) Hendriks, M. M. W. B.; Eeuwijk, F. A.; Jellema, R. H.; Westerhuis, J. A.; Reijmers, T. H.; Hoefsloot, H. C. J.; Smilde, A. K. Trends Anal. Chem. 2011, 30, 1685-1698. 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 24

(7) Boccard, J.; Veuthey, J. L.; Rudaz, S. J. Sep. Sci. 2010, 33, 290-304.

(8) Wei, R.; Li, G. D.; Seymour, A. B. Anal. Chem. 2010, 82, 5527–5533.

(9) Griffiths, W. J.; Koal, T.; Wang, Y.; Kohl, M.; Enot, D. P.; Deigner, H. P. Angew. Chem., Int. Ed. 2010, 49, 5426-5445.

(10) Kim, K. Y.; Joo, H. J.; Kwon, H. W.; Kim, H.; Hancock, W. S.; Paik, Y. K. Anal. Chem. 2013, 85, 2681-2688.

(11) Yogi, K.; Oshiro, N.; Inafuku, Y.; Hirama, M.; Yasumoto, T. Anal. Chem. 2011, 83, 8886-8891.

(12) Boyd, R. K.; Basic, C.; Bethem, R. A. Trace Quantitative Analysis by Mass Spectrometry; John Wiley & Sons Ltd: Chichester, West Sussex, U.K., 2008.

(13) Vrhovsek, U.; Masuero, D.; Gasperotti, M.; Franceschi, P.; Caputi, L.; Viola, R.; Mattivi, F. J. Agric. Food Chem. 2012, 60, 8831-8840.

(14) Guo, B.; Chen, B.; Liu, A.; Zhu, W.; Yao, S. Curr. Drug Metab. 2012, 13, 1226-1243.

(15) Yuan, M.; Breitkopf, S. B.; Yang, X.; Asara, J. M. Nat. Protoc. 2012, 7, 872-881.

(16) Schmerler, D.; Neugebauer, S.; Ludewig, K.; Bremer-Streck, S.; Brunkhorst, F. M.; Kiehntopf, M. J. Lipid Res. 2012, 53, 1369-1375.

16


Page 17 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(17) Ubhi, B. K.; Cheng, K. K.; Dong, J.; Janowitz, T.; Jodrell, D.; Tal-Singer, R.; MacNee, W.; Lomas, D. A.; Riley, J. H.; Griffin, J. L.; Connor, S. C. Molecular bioSystems 2012, 8, 3125-3133.

(18) Li, Y.; Ruan, Q.; Li, Y.; Ye, G.; Lu, X.; Lin, X.; Xu, G. J. Chromatogr., A 2012, 1255, 228-236.

(19) Chen, S.; Kong, H.; Lu, X.; Li, Y.; Yin, P.; Zeng, Z.; Xu, G. Anal. Chem. 2013, 85, 8326-8333.

(20) Ye, G.; Liu, Y.; Yin, P.; Zeng, Z.; Huang, Q.; Kong, H.; Lu, X.; Zhong, L.; Zhang, Z.; Xu, G. J. Proteome Res. 2014, 13, 1994-2004.

(21) Tautenhahn, R.; Patti, G. J.; Kalisiak, E.; Miyamoto, T.; Schmidt, M.; Lo, F. Y.; McBee, J.; Baliga, N. S.; Siuzdak, G. Anal. Chem. 2011, 83, 696-700.

(22) Dane, A. D.; Hendriks, M. M.; Reijmers, T. H.; Harms, A. C.; Troost, J.; Vreeken, R. J.; Boomsma, D. I.; van Duijn, C. M.; Slagboom, E. P.; Hankemeier, T. Anal. Chem. 2014, 86, 4110-4114.

(23) Zhou, R.; Tseng, C. L.; Huan, T.; Li, L. Anal. Chem. 2014, 86, 4675-4679.

(24) Yin, P.; Peter, A.; Franken, H.; Zhao, X.; Neukamm, S. S.; Rosenbaum, L.; Lucio, M.; Zell, A.; Haring, H. U.; Xu, G.; Lehmann, R. Clin. Chem. 2013, 59, 833-845.

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 24

(25) Zeng, Z.; Liu, X.; Dai, W.; Yin, P.; Zhou, L.; Huang, Q.; Lin, X.; Xu, G. Anal. Chem. 2014, 86, 3793-3800.

(26) Fernandez-Albert, F.; Llorach, R.; Andres-Lacueva, C.; Perera-Lluna, A. Anal. Chem. 2014, 86, 2320-2325.

(27) Werner, E.; Heilier, J. F.; Ducruix, C.; Ezan, E.; Junot, C.; Tabet, J. C. J. Chromatogr., B 2008, 871, 143-163.

(28) Zhou, L.; Ding, L.; Yin, P.; Lu, X.; Wang, X.; Niu, J.; Gao, P.; Xu, G. J. Proteome Res. 2012, 11, 5433-5442.

(29) Tan, Y.; Yin, P.; Tang, L.; Xing, W.; Huang, Q.; Cao, D.; Zhao, X.; Wang, W.; Lu, X.; Xu, Z.; Wang, H.; Xu, G. Mol. Cell Proteomics 2012, 11, M111 010694.

(30) Zhou, L.; Wang, Q.; Yin, P.; Xing, W.; Wu, Z.; Chen, S.; Lu, X.; Zhang, Y.; Lin, X.; Xu, G. Anal. Bioanal. Chem. 2012, 403, 203-213.

(31) Barber M.N.; Risis S.; Yang C.; Meikle P.J; Staples M.; Febbraio M.A.; C.R.., B. PloS one 2012, 7, e41456.

(32) Watanabe, M.; Houten, S. M.; Mataki, C.; Christoffolete, M. A.; Kim, B. W.; Sato, H.; Messaddeq, N.; Harney, J. W.; Ezaki, O.; Kodama, T.; Schoonjans, K.; Bianco, A. C.; Auwerx, J. Nature 2006, 439, 484-489.

18


Page 19 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(33) Chen, J.; Zhang, X.; Cao, R.; Lu, X.; Zhao, S.; Fekete, A.; Huang, Q.; Schmitt-Kopplin, P.; Wang, Y.; Xu, Z.; Wan, X.; Wu, X.; Zhao, N.; Xu, C.; Xu, G. J. Proteome Res. 2011, 10, 2625-2632.

19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 24

Figure Caption Figure 1. The workflow of pseudo-targeted metabolomics and MRM ion pairs selection. Figure 2. Comparison of metabolites coverage of non-targeted and pseudo-targeted metabolomics. (a) The total ion chromatorgraphys (TIC) of pooled serum samples with different CE voltages (20 V, 40 V, 60 V). (b) The Venny diagram illustractes the results of the extracted ions from MarkerView and MRM-Ion Pair Finder software, respectively. (c) and (d) The 2D-plot and 3D-plot show the distribution of ions extracted by MarkerView and MRM-Ion Pair Finder softwares, respectively. Figure 3. (a) Z-score plot of differential metabolites between healthy controls and patients with HCC. (b) Alteration of an unidentified differential metabolite ion No.749. (c) and (d) show the extracted MS1 and MS2 spectra of the ion No.749, respectively. The ion No.1354 (precursor ion=583.26Da, product ion=297 Da, tR=10.7 min) was further identified as biliverdin by using our previously developed method. Abbreviations: LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; PC, phosphatidylcholine; CAN, carnitine; TCA, taurocholic acid; GCA, glycocholic acid; TCDCA, taurochenodeoxycholate; GUDCA, glycochenodeoxycholate GCDCA, glycochenodeoxycholate.

20


Page 21 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. The workflow of pseudo-targeted metabolomics and MRM ion pairs selection.

21



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 24

Figure 2. Comparison of metabolites coverage of non-targeted and pseudo-targeted metabolomics. (a) The total ion chromatorgraphys (TIC) of pooled serum samples with different CE voltages (20 V, 40 V, 60 V). (b) The Venny diagram illustractes the results of the extracted ions from MarkerView and MRM-Ion Pair Finder software, respectively. (c) and (d) The 2D-plot and 3D-plot show the distribution of ions extracted by MarkerView and MRM-Ion Pair Finder softwares, respectively.

22


Page 23 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. (a) Z-score plot of differential metabolites between healthy controls and patients with HCC. (b) Alteration of an unidentified differential metabolite ion No.749. (c) and (d) show the extracted MS1 and MS2 spectra of the ion No.749, respectively. The ion No.1354 (precursor ion=583.26Da, product ion=297 Da, tR=10.7 min) was further identified as biliverdin by using our previously developed method. Abbreviations: LPC, lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; PC, phosphatidylcholine; CAN, carnitine; TCA, taurocholic acid; GCA, glycocholic acid; TCDCA, taurochenodeoxycholate; GUDCA, glycochenodeoxycholate GCDCA, glycochenodeoxycholate.

23



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 24

For TOC only.

24


Multiple Reaction Monitoring-Ion Pair Finder: A ... - ACS Publications

Recommend Documents